Disasters

Bangladesh Delta Floodings: An Integrated Framework for Social Adaptation to Cyclone Surges

Region Asia


Summary

Trialled in the Bengal Delta in Bangladesh, BanD-SOS aims to establish a pre-operational service for forecasting cyclonic flooding and the associated societal risk. The main component of this service is the forecast of flooding hazards in real time. In addition, coastal vulnerability indices at the sub-district level are disseminated with flood forecasts to provide a quick view of the vulnerability of the exposed population.

BanD-SOS marks the first-ever deployment of an operational coastal flood forecasting platform capable of predicting cyclonic flooding in the Bengal Delta. The platform is currently operated by the Bangladesh Water Development Board (BWDB), the governmental water management organisation in Bangladesh and a partner public authority of the project. The dissemination platform is accessible via an immediately useful web portal, accessible to all users, in real time (https://bandsos.github.io). The system was deployed and has been running continuously since November 2022. It has been put to the test in real life during several cyclones since its deployment. Among these events, Cyclone Remal in May 2024 is noteworthy, which severely impacted the coastal region of Bangladesh. The forecast provided by BanD-SOS was used by BWDB in better preparing for the management of this disaster with localised flood forecasts. The BanD-SOS service is managed by local engineers at BWDB and is being integrated into the larger BWDB forecasting capabilities as an essential tool. BanD-SOS is in its 4th year of operation and continues to operate for the ongoing 2026 cyclone season.

In early 2026, the platform migrated to a more reliable computing environment and updated to accommodate changes in upstream data sources, including NOAA global atmospheric forecasts. Future improvements by engineers at BWDB are planned, including incorporation of new bathymetric datasets in the highly dynamic Bengal delta region, as well as enhancement of the online dissemination portal to meet new local needs.

Value Chain Diagram

  1. Decide

    • Understand the impact of cyclonic flooding on vulnerable populations in the Bengal Delta to support emergency response and evacuation planning.
    • Provide authorities with actionable forecasts in near real-time to reduce risk and improve resilience.
  2. Apply

    • BanD-SOS pre-operational forecasting service integrates satellite-derived elevation data to develop a high-resolution cross-scale hydrodynamic modelling system, with real-time outputs routinely disseminated in a web-based flood forecasting platform for local authorities and other users.
    • Post project completion, water level forecasts generated from BanD-SOS are adopted as the oceanic boundary condition for the existing national river forecasting platform.
    Service Evolution
  3. Data production

    • Digital Elevation Model (DEM) of the intertidal zone along the coastline.
    • Flood hazard forecasts combined with socio-spatial vulnerability mapping
    Ground Segment Evolution
  4. Sense

    • Sentinel 2A and 2B multispectral imaging
    • Bangladesh Bureau of Statistics databases (SVRS, HIES)
    Space Segment Evolution

Decision-Making Context and Needs

The Bengal Delta, straddling Bangladesh and India, is hit by a major tropical cyclone every 3 years on average. Their occurrence is relatively moderate compared to other tropical regions, representing only 5% of global cyclones. The impact of these events is, however, major, representing 50% of the cyclone victims recorded globally. This is due to the very low topography of the delta above sea level (less than 5 metres) and the fragility of coastal defense structures, combined with the high density of the vulnerable population (Figure 1)

Figure 1: Trajectories and intensities of major cyclones that have struck the Bengal Delta over the past five decades, with the number of victims recorded for each. (Attribution: © Jamal Uddin Khan (LIENSs, La Rochelle))

The exposure of the Bengal Delta to cyclonic flooding is increasing in the current context of sea level rise due to climate change.

The objective of the BanD-SOS project was to develop a pre-operational flood forecasting system whose function is to:

  • provide in real time the basic information necessary for the implementation of evacuation and protection operations when a hurricane event occurs;
  • assist in the development of public policies for adaptation to the flood hazard necessary to ensure the resilience of the Delta population in the medium to long term.

To this end, the BanD-SOS service consisted of pre-operationalizing a numerical forecasting platform for cyclonic flooding over Bengal, for which we have completed the proof of concept (Khan et al., 2021). The BanD-SOS service ensures the prediction of the flood hazard, combined with spatialised information of the socio-spatial vulnerability of the exposed territories.

Figure 2: Coastal defenses installed along the coastline of Kuakata (South Bangladesh) (Attribution: © Jamal Uddin Khan (LIENSs, La Rochelle))

The main user of the first operational service available is the Bangladesh Flood Forecasting and Warning Center (FFWC).

Earth Observation Data Products and Information

The BanD-SOS platform uses the SCHISM hydrodynamic model coupled with the WWM wave model, which simulates the temporal evolution of sea level and waves every 5 minutes on a scale of a hundred metres. Considering the interaction of the tide, the cyclonic effect, and river inputs from the mainland, it reproduces sea level variability across the entire coastal ocean-delta-river continuum. It is particularly well suited to capturing the extreme water levels that develop at the peak of cyclonic episodes.

Spatial remote sensing is at the heart of the BanD-SOS system. Thanks to an archive of high-resolution Sentinel-2 satellite images, the project team has been able to calculate a Digital Elevation Model (DEM) of the area along the coastline between high and low tide, known as the “intertidal zone.” This is a key element of the BanD-SOS platform, as cyclonic inundation dynamics depend on the topography of the shallow shorefront and the intertidal zone. This is particularly important for the Bengal Delta with vast, gently sloping beaches and typically not observed by topographic nor hydrographic surveys.

Underlying Observations and Data Sources

The BanD-SOS system relies on a combination of satellite and ground-based data to produce accurate forecasts of cyclonic flooding. The main satellite inputs come from Sentinel-2A and -2B multi-spectral imaging, which provide high-resolution imagery necessary to calculate the Digital Elevation Model (DEM) of the intertidal zones along the Bengal Delta coastline.

Figure 3: Flood in Bangladesh (Attribution: © Jamal Uddin Khan (LIENSs, La Rochelle))

In addition to satellite data, BanD-SOS uses information from the Bangladesh Bureau of Statistics, including the Sample Vital Registration System (SVRS) and the Household Income and Expenditure Surveys (HIES). These datasets provide socio-economic and population information that is essential for mapping the vulnerability of populations exposed to flooding. Together, these sources enable BanD-SOS to couple hydrodynamic forecasts with spatialised vulnerability information, providing actionable insights for both emergency response and long-term planning.

The system is continuously operated and maintained by BWDB engineers, ensuring sustainability and reliability of the forecasts. The recent migration to a more robust computing environment and incorporation of updated upstream data sources (e.g., NOAA global atmospheric forecasts) further strengthen the operational quality of the service.

Impacts and Benefits

The BanD-SOS project established a pre-operational production chain for coastal flooding prediction, accessible on demand via a web interface for non-specialist end users. The platform is fully open-source, hosted on GitHub Pages, and allows users to view forecasts for the next 14 days as well as replay past events. Results are fully public and easily interpretable by non-experts, enabling authorities and other stakeholders to take timely action.

Figure 4: Bangladesh, July 16th, 2020. Storm surges caused by more frequent and powerful cyclones push walls of water 80 to 100 km along the delta rivers. (Attribution: © Moniruzzaman Sazal / Climate Visuals Countdown)

During Cyclone Remal, which formed unusually fast in 3 days and lasted 36 hours, more than 800,000 Bangladeshis evacuated the coast. Approximately 4.6 million people were affected, 40,000 homes were destroyed, and 133,000 were damaged. The BanD-SOS forecasts closely matched observed flooding, capturing three successive peaks of maximum water levels of 2–3 meters above high tide.

According to the Flood Forecasting and Warning Center (FFWC), BanD-SOS supported field officials in preparing for inundation and potential infrastructure failures. Future improvements are planned, including adjusting forecasts with local mean sea level (MSL) observations to increase accuracy. Overall, the project demonstrates the effectiveness of an operational, real-time flood forecasting system in saving lives, informing emergency response, and supporting the resilience of vulnerable populations.

The demonstration of the replication of this pre-operational chain will be realized with the porting of the solution to the Amazon Delta, highlighting the potential for broader application beyond the Bengal Delta.